System and method for diagnosing pathologic heart conditions

• US 7,001,338 B2
• Filed: 03/24/2004
• Issued: 02/21/2006
• Est. Priority Date: 03/26/2003
• Status: Active Grant

• Current Assignee: Johns Hopkins University
• Original Assignee: Johns Hopkins University
• Inventors: Hayek, Carleton S., Thompson, W. Reid
• Primary Examiner(s): Pezzuto, Robert E.
• Assistant Examiner(s): Gedeon, Brian T.
• Application Number:
  US10/808,196
• Publication Number:
  US 20040193067A1
• Time in Patent Office: 699 Days
• Field of Search: 600/528, 600/512, 600/514, 600/508
• US Class Current: 600/528
• CPC Class Codes: A61B 5/349 Detecting specific parameters of the electrocardiograph cycle
• CPC Class Codes: A61B 5/726 using Wavelet transforms
• CPC Class Codes: A61B 7/00 Instruments for auscultation
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R

Abstract

A method of diagnosing pathologic heart conditions in which a time series of heart sounds is filtered and parsed into a sequence of individual heart cycles. A systolic interval as well as systolic sub-intervals are identified for each heart cycle. The systolic intervals and ECG peaks are then digitally filtered to optimize for click detection. For each heartcycle, systole time limits are determined, a time series of the transform at specific wavelet scales are input to a Neyman-Pearson “constant false alarm rate” (CFAR) detector to identify anomalously high wavelet coefficients, and a vector of detections vs. time is created. The series of anomalously high detections (one series for each heart cycle) are then assembled into a matrix and convolved with an averaging vector yielding detection statistics across heart cycles and time intervals consistent with an observed spread of click occurrence times. A click score is then determined as the maximum element of the vector formed by the median wavelet coefficient amplitude across heart cycles squared at each time sample multiplied element-wise by the vector formed by the sum across heart cycles of the number of detections at each time sample. The click score is compared to a threshold value set by a desired probability of detection vs. a probability of false alarm tradeoff. If the click score is less than the threshold then a “no click” indicator is displayed. If the click score is greater than the threshold then a “click present” indicator is displayed.

Claims

1. A method of diagnosing pathologic heart conditions comprising:

parsing a time series of heart sounds into a sequence of individual heart cycles;

determining the systolic interval of the heart cycles;

identifying a subinterval for each systolic interval of the heart cycles;

filtering the time series of heart sounds using a time-frequency transformation;

identifying anomalously high wavelet coefficients using a constant false alarm rate (CFAR) detector;

assembling the series of anomalously high detections into a matrix;

convolving the matrix with an averaging vector, said vector'"'"'s length derived from the expected time spread of a click occurrence, the convolution yielding detection statistics across heart cycles and time intervals consistent with an observed spread of click occurrence times;

calculating a click score as the maximum element of a vector formed by the median wavelet coefficient amplitude across heart cycles squared at each time sample multiplied element-wise by a vector formed by the sum across heart cycles of the number of detections at each time sample; and

comparing the click score to a threshold level in order to distinguish between a normal heart and a pathologic heart.

2. The method of claim 1 further comprising ranking the click score relative to other click scores in a database to establish its standing amongst a population of other click scores of known pathologic and non-pathologic hearts.

3. The method of claim 1 wherein parsing the time series of heart sounds into a sequence of individual heart cycles uses electro-cardiogram (ECG) data to transform a time series of heart sounds into a sequence of individual heart cycles.

4. The method of claim 1 wherein parsing the time series of heart sounds into a sequence of individual heart cycles uses acoustic heart sounds obtained directly from a patient to transform a time series of heart sounds into a sequence of individual heart cycles.

5. The method of claim 1 wherein determining the systolic interval of the heart cycles is achieved by identifying pulses on an electro-cardiogram (ECG).

6. The method of claim 1 wherein determining the systolic interval of the heart cycles is achieved by acoustically locating a first and a second heart sound using a bandpass filter, said bandpass filter applied to the time series of heart sounds.

7. The method of claim 1 wherein the systolic sub-interval is centered in the systolic interval.

8. The method of claim 1 wherein the systolic sub-interval is centered in systole and spans half of the systolic interval.

9. The method of claim 1 wherein filtering the time series of heart sounds using a time-frequency transformation is implemented by a second order coiflet continuous wavelet transform (CWT).

10. The method of claim 1 wherein filtering the time series of heart sounds using a time-frequency transformation is implemented by a Fourier transform.

11. A system for diagnosing pathologic heart conditions comprising:

means for parsing a time series of heart sounds into a sequence of individual heart cycles;

means for determining the systolic interval of the heart cycles;

means for identifying a subinterval for each systolic interval of the heart cycles;

means for filtering the time series of heart sounds using a time-frequency transformation;

means for identifying anomalously high wavelet coefficients using a constant false alarm rate (CFAR) detector;

means for assembling the series of anomalously high detections into a matrix;

means for convolving the matrix with an averaging vector, said vector'"'"'s length derived from the expected time spread of a click occurrence, the convolution yielding detection statistics across heart cycles and time intervals consistent with an observed spread of click occurrence times;

means for calculating a click score as the maximum element of a vector formed by the median wavelet coefficient amplitude across heart cycles squared at each time sample multiplied element-wise by a vector formed by the sum across heart cycles of the number of detections at each time sample;

means for comparing the click score to a threshold level in order to distinguish between a normal heart and a pathologic heart; and

means for ranking the click score relative to others in a database to establish its standing amongst a population of other scores of known pathologic and non-pathologic cases.

12. The system of claim 11 further comprising means for ranking the click score relative to other click scores in a database to establish its standing amongst a population of other click scores of known pathologic and non-pathologic hearts.

13. The system of claim 11 wherein the means for parsing the time series of heart sounds into a sequence of individual heart cycles uses electrocardiogram (ECG) data to transform a time series of heart sounds into a sequence of individual heart cycles.

14. The system of claim 11 wherein the means for parsing the time series of heart sounds into a sequence of individual heart cycles uses acoustic heart sounds obtained directly from a patient to transform a time series of heart sounds into a sequence of individual heart cycles.

15. The system of claim 11 wherein the means for determining the systolic interval of the heart cycles is achieved by identifying pulses on an electro-cardiogram (ECG).

16. The system of claim 11 wherein the means for determining the systolic interval of the heart cycles is achieved by acoustically locating a first and a second heart sound using a bandpass filter, said bandpass filter applied to the time series of heart sounds.

17. The system of claim 11 wherein the systolic sub-interval is centered in the systolic interval.

18. The system of claim 11 wherein the systolic sub-interval is centered in systole and spans half of the systolic interval.

19. The system of claim 11 wherein the means for filtering the time series of heart sounds using a time-frequency transformation is implemented by a second order coiflet continuous wavelet transform (CWT).

20. The system of claim 11 wherein the means for filtering the time series of heart sounds using a time-frequency transformation is implemented by a Fourier transform.

21. A computer readable medium comprising a computer program for diagnosing pathologic heart conditions, the computer program product comprising:

computer program code for parsing a time series of heart sounds into a sequence of individual heart cycles;

computer program code for determining the systolic interval of the heart cycles;

computer program code for identifying a subinterval for each systolic interval of the heart cycles;

computer program code for filtering the time series of heart sounds using a time-frequency transformation;

computer program code for identifying anomalously high wavelet coefficients using a constant false alarm rate (CFAR) detector;

computer program code for assembling the series of anomalously high detections into a matrix;

computer program code for convolving the matrix with an averaging vector, said vector'"'"'s length derived from the expected time spread of a click occurrence, the convolution yielding detection statistics across heart cycles and time intervals consistent with an observed spread of click occurrence times;

computer program code for calculating a click score as the maximum element of a vector formed by the median wavelet coefficient amplitude across heart cycles squared at each time sample multiplied element-wise by a vector formed by the sum across heart cycles of the number of detections at each time sample;

computer program code for comparing the click score to a threshold level in order to distinguish between a normal heart and a pathologic heart; and

computer program code for ranking the click score relative to others in a database to establish its standing amongst a population of other scores of known pathologic and non-pathologic cases.

22. The computer program of claim 21 further comprising computer program code for ranking the click score relative to other click scores in a database to establish its standing amongst a population of other click scores of known pathologic and non-pathologic hearts.

23. The computer program of claim 21 wherein the computer program code for parsing the time series of heart sounds into a sequence of individual heart cycles uses electro-cardiogram (ECG) data to transform a time series of heart sounds into a sequence of individual heart cycles.

24. The computer program of claim 21 wherein the computer program code for parsing the time series of heart sounds into a sequence of individual heart cycles uses acoustic heart sounds obtained directly from a patient to transform a time series of heart sounds into a sequence of individual heart cycles.

25. The computer program of claim 21 wherein the computer program code for determining the systolic interval of the heart cycles is achieved by identifying pulses on an electrocardiogram (ECG).

26. The computer program of claim 21 wherein the computer program code for determining the systolic interval of the heart cycles is achieved by acoustically locating a first and a second heart sound using a bandpass filter, said bandpass filter applied to the time series of heart sounds.

27. The computer program of claim 21 wherein the systolic sub-interval is centered in the systolic interval.

28. The computer program of claim 21 wherein the systolic sub-interval is centered in systole and spans half of the systolic interval.

29. The computer program of claim 21 wherein the computer program code for filtering the time series of heart sounds using a time-frequency transformation is implemented by a second order coiflet continuous wavelet transform (CWT).

30. The computer program of claim 21 wherein the computer program code for filtering the time series of heart sounds using a time-frequency transformation is implemented by a Fourier transform.